วีดีโอ

Ehhh... it's close enough

Because CPUs did not attend elementary school

At the point of input conversion Python makes a mistake, encoding 0.1 and 0.2 as binary floating point numbers, while the input is clearly given as decimal floating point. This has nothing to do with CPU or floating point or whatever. It's a software mistake, one of many unfixed Python mistakes.

C++ And C

4,(max calculater amount of 0s)- 399999999999 = 0 what??? why does it say 1000000000 proof its ture

I develop with Delphi, when someone create a condition with floats its very important use Round with a significant decimal places, because if you compare two floats sometimes the processor can't do properly exactly by the reason explained in the vídeo, its like compare two variables "(0.12 = 0.12) = false" but real value of one is "0.1199573638" and the program execute a wrong instruction

how c++ can do it while python made from c++?

Just calculate on your own and print it.

THAT MUSIC AT THE START NOSTALGIA

10001 = 11?

Skill issue

But nowdays 64gb of ram is almost standard, why doesnt "real data point" is used?

The Orange Game song at the beginning. Nostalgia!

julia> 1//10+2//10 3//10

echo 0.1+0.2 | bc -l .3

Silicon errors. Silicon inpuriries. Your mom.

2:15 ....ahhh....it hurts....you are monster!.....see u in hell dude

0:56 me whos still in hs 💀

"point floating"? you mean FLOATING POINT- there's no such thing as "point floating".

Just a little correction on the title, it's correctly written as "why does the google calculator fail at math". You can change the title, please do it.

So what you refer to simple AI like chatbot, I’d say that’s not even AI.. the word AI is overly used these days

"pyton"

Is this also why multiplying two large, but finite numbers end up as “Infinity?”

in lua 0.1 + 0.2 somehow gave 0.3 in the online compiler

at least the apple calculator (accessable via app store for iphones, or cmd+space for ipads/macs) has it right, right?

In my Google Calculator app, it works fine! I tested it and it worked perfectly...

floating point duh

2:16 Noooooo

PRINT 0.1+0.2 prints out 0.3 on my ZX Spectrum.

Their measurement converter also works incorrectly. If you write integer, it works fine, but if you type float value, it adjusts formula to have same answer as if you wrote integer in it. (Example: you write 1.3 meters but it will give you 100cm instead of 130cm)

This is one case where bing is actually better than google

Same reason on why you can't enter more than 15 digits on an android calculator

So how to fix it?

lmao i tried the 4000000000000000001-4000000000000000000 thing and it's really sad that it equaled 0

People here are complaining about the fact that you made a video about floating point precision, I think the major issue here is that the programmers who built Google's calculator used floats for it, most languages have builtin BigNumber support, which is a structure that allows you to do math at arbitrary precision. As you said, we don't have infinite memory, but we don't need to go too far beyond the limits of a modern computer with any calculation you can type into the search box. And if the calculations happen on the client side, they could be run async and have a loading spinner in case the numbers are really large

The built in windows calculator seems to handle it just fine so I think it is fair to say the google calculator has failed at its task.

another floating point video... 💤💤😴

you made my google home start listening by saying "...okay google, are you doing arlight??"

floating point precision is common knowledge why am i seeing so many videos about it

Because programmer is ugly

what's the song on the first part

In case someone wants to know more about this, just read IEEE754. I had to learn it the hard way. But to be honest, in my opinion this detail is relatively unimportant in my daily programming business. In fact, I was programming for several years before I even learned about it. Sometimes I was wondering about odd results but none of them was important enough to seriously mess with the things I was trying to achieve.

another floating point video DJKFHSJKHFLKJDHFJSD

nvm i tried it and it also gave 0

it probably reached 64 bit integer × 64 bit integer

Heyo, I messed up the numbers a little bit in the video, noticed it only on the editing stage :(

What if math is wrong? 🤔

Isn't adding the job of the ALU and not the CPU?

Please call it floating point, not point floating

Bro erase that before AI takes control, otherwise they will find you hehehee